Abstract
Discovery of biomarkers using serum proteomic patterns is currently one of the most attractive interdisciplinary research areas in computational life science. This new proteomic approach has the clinical significance in being able to detect disease in its early stages and to develop new drugs for disease treatment and prevention. This paper introduces a novel pattern classification strategy for identifying protein biomarkers using mass spectrometry data of blood samples collected from patients in emergency department monitored for major adverse cardiac events within six months. We applied the theory of geostatistics and a kriging error matching scheme for identifying protein biomarkers that are able to provide an average classification rate superior to other current methods. The proposed strategy is very promising, as a general computational bioinformatic model for proteomic-pattern based biomarker discovery.